sahilmaniyar888/isl-healthcare-instruction-corpus

首个开放的印度手语（ISL）医疗指令调优基准数据集，专注于医疗保健领域。该数据集包含1174个样本，分为1074个词汇符号和100个上下文指令，涵盖了从解剖学到疫苗接种等多个医疗专业领域。每个样本包含英语和印地语的自然语言文本，以及印度手语的注释序列。数据集旨在支持印度手语在医疗领域的应用和研究，特别是在低资源环境下的适应性数据挑战。The first open Indian Sign Language (ISL) instruction-tuning benchmark focused on healthcare. The dataset contains 1174 samples, divided into 1074 lexical signs and 100 contextual instructions, covering various medical specialties from anatomy to vaccination. Each sample includes natural language text in English and Hindi, as well as gloss sequences for Indian Sign Language. The dataset is designed to support the application and research of ISL in the medical field, particularly for adaptive data challenges in low-resource environments.